Small-sample inference with spatial HAC estimators

This paper examines the small-sample performance of spatial HAC (SHAC) estimators of the standard errors on parameters. We find that, in small to moderately-sized datasets, the use of HAC estimators may be recommended only with a relatively large degree of cross-sectional interdependence.     

On justifying a minimum welfare state

Since anarchy is not viable, limited government is the best that the realistic libertarian can hope for. But limited government will itself always be threatened by an inherent tendency to transgress its limits. In modern western societies the regulatory and redistributive welfare state is the major threat to a constitution of liberty. However, a “minimum welfare state” which redistributes personal income among its citizens may comply with the same principles of individual liberty and the rule of law that are embodied in the protective state. Since any state, including the minimal state, necessarily incorporates regulation and redistribution and thus is a welfare state of sorts the non-anarchist liberal should turn against welfare state privileges rather than against redistribution and regulation per se. He may even have good reason to go beyond the minimal state to found a “minimum welfare state” if this is instrumental in securing liberty under the rule of law. I am grateful to the Center for the Study of Public Choice, George Mason University for hospitality both during the period in which this paper was written and on other occasions. I am deeply indebted to the people at the Center for their criticisms and comments. As far as this paper is concerned Geoffrey Brennan's and Richard Wagner's comments were particularly helpful. I should also like to acknowledge helpful oral comments from Kevin Mulligan and Philip van Parijs, who of course is much more of an expert on demogrant schemes than I am. I also wish to thank two anonymous referees for their valuable suggestions. The general caveat applies.